Spaces:
Running
Running
Se agrego gitignore y se modifico segun recomendaciones el app.yp
Browse files- .gitignore +1 -0
- app.py +22 -38
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
.vscode
|
app.py
CHANGED
@@ -1,63 +1,47 @@
|
|
1 |
import gradio as gr
|
2 |
from huggingface_hub import InferenceClient
|
3 |
|
4 |
-
"""
|
5 |
-
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
|
6 |
-
"""
|
7 |
client = InferenceClient("jhangmez/CHATPRG-v0.2.1-Meta-Llama-3.1-8B-bnb-4bit-q4_k_m")
|
8 |
|
|
|
|
|
9 |
|
10 |
-
def respond(
|
11 |
-
|
12 |
-
history: list[tuple[str, str]],
|
13 |
-
system_message,
|
14 |
-
max_tokens,
|
15 |
-
temperature,
|
16 |
-
top_p,
|
17 |
-
):
|
18 |
-
messages = [{"role": "system", "content": system_message}]
|
19 |
|
20 |
-
for
|
21 |
-
if
|
22 |
-
|
23 |
-
if
|
24 |
-
|
25 |
|
26 |
-
|
|
|
27 |
|
28 |
-
|
29 |
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
temperature=temperature,
|
35 |
top_p=top_p,
|
|
|
36 |
):
|
37 |
-
|
38 |
-
|
39 |
-
response += token
|
40 |
-
yield response
|
41 |
|
42 |
-
"""
|
43 |
-
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
|
44 |
-
"""
|
45 |
demo = gr.ChatInterface(
|
46 |
respond,
|
47 |
additional_inputs=[
|
48 |
gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
|
49 |
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
|
50 |
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
|
51 |
-
gr.Slider(
|
52 |
-
minimum=0.1,
|
53 |
-
maximum=1.0,
|
54 |
-
value=0.95,
|
55 |
-
step=0.05,
|
56 |
-
label="Top-p (nucleus sampling)",
|
57 |
-
),
|
58 |
],
|
|
|
|
|
59 |
)
|
60 |
|
61 |
-
|
62 |
if __name__ == "__main__":
|
63 |
demo.launch()
|
|
|
1 |
import gradio as gr
|
2 |
from huggingface_hub import InferenceClient
|
3 |
|
|
|
|
|
|
|
4 |
client = InferenceClient("jhangmez/CHATPRG-v0.2.1-Meta-Llama-3.1-8B-bnb-4bit-q4_k_m")
|
5 |
|
6 |
+
def format_message(role, content):
|
7 |
+
return f"<|im_start|>{role} {content}<|im_end|> "
|
8 |
|
9 |
+
def respond(message, history, system_message, max_tokens, temperature, top_p):
|
10 |
+
formatted_messages = [format_message("system", system_message)]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
|
12 |
+
for human, assistant in history:
|
13 |
+
if human:
|
14 |
+
formatted_messages.append(format_message("user", human))
|
15 |
+
if assistant:
|
16 |
+
formatted_messages.append(format_message("assistant", assistant))
|
17 |
|
18 |
+
formatted_messages.append(format_message("user", message))
|
19 |
+
formatted_messages.append("<|im_start|>assistant ")
|
20 |
|
21 |
+
full_prompt = "".join(formatted_messages)
|
22 |
|
23 |
+
response = ""
|
24 |
+
for chunk in client.text_generation(
|
25 |
+
full_prompt,
|
26 |
+
max_new_tokens=max_tokens,
|
27 |
temperature=temperature,
|
28 |
top_p=top_p,
|
29 |
+
stream=True,
|
30 |
):
|
31 |
+
response += chunk
|
32 |
+
yield response.strip()
|
|
|
|
|
33 |
|
|
|
|
|
|
|
34 |
demo = gr.ChatInterface(
|
35 |
respond,
|
36 |
additional_inputs=[
|
37 |
gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
|
38 |
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
|
39 |
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
|
40 |
+
gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
|
|
|
|
|
|
|
|
|
|
|
|
|
41 |
],
|
42 |
+
title="Meta Llama 3.1 8B Chatbot",
|
43 |
+
description="A chatbot powered by Meta Llama 3.1 8B model."
|
44 |
)
|
45 |
|
|
|
46 |
if __name__ == "__main__":
|
47 |
demo.launch()
|